https://github.com/Microsoft/CNTK

sort by:
Revision Author Date Message Commit Date
381506d updating baseline.txt 08 May 2017, 18:48:13 UTC
a527f82 move linux java tests 08 May 2017, 18:37:58 UTC
3485325 moving test location, expanding tests to gpu, fixing comments 08 May 2017, 17:48:21 UTC
e326497 Merge branch 'marhamil/java-bindings' of https://github.com/Microsoft/cntk into marhamil/java-bindings 08 May 2017, 15:01:55 UTC
a49ec85 changing test location 08 May 2017, 15:01:51 UTC
59857a1 use List interface instead of ArrayList implementation 08 May 2017, 14:55:42 UTC
5051c65 DRY in Main.java 08 May 2017, 14:42:40 UTC
d8f6b3f ignore DeviceDescriptorVector size constructor 08 May 2017, 14:42:12 UTC
6232ec4 typo 08 May 2017, 13:57:48 UTC
9480d8c Java Bindings adding some doc use $$ add comments automatically loadLibrary in java semicolon removing .iml add more dependencies move java static block code to cntk_java.i remove also class files. Adding .java files to jar 08 May 2017, 02:21:24 UTC
190dc1b CNTK v2 library: Fixed a bug in shape inference for element operations when eiter of the operands has free static axes. 07 May 2017, 18:28:37 UTC
96ed208 CNTk v2 library: Add support for free static axes in UserDefinedFunctions. 07 May 2017, 09:00:14 UTC
9fd84c8 CNTK v2 library: Fix handling of scalar loss functions in Trainer. 07 May 2017, 00:23:36 UTC
74ad63f CNTK v2 library: Fix evaluation of Function when a Constant is specified as one of the requested Outputs. 06 May 2017, 23:22:07 UTC
6e6e109 CNTK v2 library: Fix a bug in shape inference for Slice, when one or more inputs have InferredDimension along the splice axis. 06 May 2017, 19:57:26 UTC
eea6943 Integrate kedeng/sparseDist2 into master 06 May 2017, 11:03:42 UTC
d9e12fe Integrate walkerning/adamax into master 06 May 2017, 05:05:44 UTC
a5703c7 Allow sparse gradient to be accumulated in distributed training 06 May 2017, 00:01:40 UTC
2bc980e Small fix for CR. 05 May 2017, 21:46:11 UTC
a5be553 Fix for 1D convolution. 05 May 2017, 20:17:43 UTC
3866fd4 Update test baseline 05 May 2017, 18:14:08 UTC
bdb4d13 Merge commit '13e019b' into walkerning/adamax 05 May 2017, 17:10:50 UTC
3cbc444 bindings/python/setup.py: add missing dependencies for incremental build 05 May 2017, 16:51:04 UTC
dc12a39 Integrate t-mofun/dqn into master 05 May 2017, 16:43:19 UTC
fb509d7 Removed unneeded python section 05 May 2017, 13:35:22 UTC
13e019b Add adamax option for Adam learner. 05 May 2017, 11:07:09 UTC
a76f253 Added README with more context about this example 05 May 2017, 10:31:45 UTC
151f325 Turn plotting to false by default 05 May 2017, 10:25:42 UTC
3a73713 WIP. adding Readme 05 May 2017, 09:27:02 UTC
53956a0 Addressing CR comments, fix typos. 05 May 2017, 09:03:08 UTC
4a2ea2e Merge branch 'lewlu/inceptionv3' of https://github.com/Microsoft/CNTK into lewlu/inceptionv3 05 May 2017, 03:06:34 UTC
25e4966 Update readme.md 05 May 2017, 03:05:59 UTC
59cd580 Address CR comments 05 May 2017, 03:05:58 UTC
7a4e3ed Remove empty line 05 May 2017, 03:05:58 UTC
9ee92f4 Update InceptionV3 05 May 2017, 03:05:58 UTC
3b93cd8 update Readme.md 05 May 2017, 03:05:58 UTC
edbc178 Update readme.md 05 May 2017, 02:59:01 UTC
b4b4ae2 Address CR comments 05 May 2017, 02:53:36 UTC
19c5373 Remove empty line 05 May 2017, 02:43:56 UTC
97e5ea9 Updating test wrapping up as a test_ function for pytest 04 May 2017, 16:25:09 UTC
c38b1d7 Merge remote-tracking branch 'origin/t-mofun/dqn' into t-mofun/dqn # Conflicts: # Examples/ReinforcementLearning/DeepQNeuralNetwork.py 04 May 2017, 16:13:30 UTC
5d28b84 Add len() operator on the memory for testing 04 May 2017, 15:57:33 UTC
7c3af5a Fix metrics writer as None and not [None] 04 May 2017, 15:57:33 UTC
7e6938b Fix wrong axis for q target computation 04 May 2017, 15:57:33 UTC
8f8bd41 Fixed import for RC 04 May 2017, 15:57:32 UTC
f7ec19c Added memory size parameter to tune memory footprint 04 May 2017, 15:57:32 UTC
711b2b7 Updated gym dependency for all Linux python versions 04 May 2017, 15:57:32 UTC
51d1969 Fix invalid file name for test discovery 04 May 2017, 15:57:32 UTC
e674601 Added copyright 04 May 2017, 15:57:32 UTC
8adf19d Gym 0.8.2 doesn't exists ... 04 May 2017, 15:57:32 UTC
da2ce24 Conda gym atari dependency 0.8.2 04 May 2017, 15:57:32 UTC
fead914 Register the RL sample in samples.json 04 May 2017, 15:57:26 UTC
37b7dee Adding reward clipping 04 May 2017, 15:55:24 UTC
6d928f1 Fix gym atari dependencies in conda.yml 04 May 2017, 15:55:21 UTC
99202bd Adding DQN test for Linux cntk35 only for the moment 04 May 2017, 15:55:20 UTC
1b81d21 Added copyright header 04 May 2017, 15:55:11 UTC
d09b89e Adressing CR comments. Typos. 04 May 2017, 15:55:11 UTC
f853b6b Computing Q targets is now part of the computation graph 04 May 2017, 15:55:11 UTC
35cd672 Addressing CR comments. Tensorboard now optional 04 May 2017, 15:55:11 UTC
ff199d0 Addressing CR comments. Reformating comments 04 May 2017, 15:55:11 UTC
10400b0 Adressing CR comments. Fix typo 04 May 2017, 15:55:11 UTC
4cd0bd7 Addressing CR comments. Removed six dependency 04 May 2017, 15:55:11 UTC
464dad2 Addressing CR comments, added Nature reference 04 May 2017, 15:55:11 UTC
7ce03dd Adressing CR comment, more explicit variable name 04 May 2017, 15:55:10 UTC
7dfeb3d Adressing CR comments for english. 04 May 2017, 15:55:10 UTC
e4c3726 More fix around explorer 04 May 2017, 15:55:10 UTC
257988f Fix wrong affine coefficient computation 04 May 2017, 15:55:10 UTC
15e743d Fix exploration decay 04 May 2017, 15:55:10 UTC
3be2dab compute_q simplification 04 May 2017, 15:55:10 UTC
0f4303e Missing update_signature call 04 May 2017, 15:55:10 UTC
2d72450 Let agent decide when it wants to start learning 04 May 2017, 15:55:10 UTC
92d3432 Specify device to run on 04 May 2017, 15:55:10 UTC
d66eea2 Fixing Trainer model and some input shapes 04 May 2017, 15:55:10 UTC
15b2697 More doc update 04 May 2017, 15:55:10 UTC
4c2e9d7 Wrong work in doc 04 May 2017, 15:55:09 UTC
a080ca4 Fix one_hot wants list of int 04 May 2017, 15:55:09 UTC
5f060ed Added comments 04 May 2017, 15:55:09 UTC
41e0791 Fix wrong tuple order for TemporalMemory.minibatch in compute_q 04 May 2017, 15:55:09 UTC
1f4dd1a Fix visualization again 04 May 2017, 15:55:09 UTC
0384426 Fix wrong method call evaluate instead of eval 04 May 2017, 15:55:09 UTC
756bfb9 More convenient visualization 04 May 2017, 15:55:09 UTC
39e4217 Fixing input shapes 04 May 2017, 15:55:09 UTC
d2c5759 Converting raw Atari environment to ale input 04 May 2017, 15:55:09 UTC
c1d4b61 Removed update_signature call. 04 May 2017, 15:55:09 UTC
43992f6 Fix appending to wrong metrics buffer 04 May 2017, 15:55:08 UTC
f66d815 Fixing Tensor meta declaration 04 May 2017, 15:55:08 UTC
b8d367d Addressing wrong tensor shapes 04 May 2017, 15:55:08 UTC
61d6cfe Adding Tensorboard visualization 04 May 2017, 15:55:08 UTC
2b2cf2d Addressing Frank CR comments 04 May 2017, 15:55:08 UTC
cad7bee Fixing typo 04 May 2017, 15:55:08 UTC
d9721b0 Adding target network update every specified intervals 04 May 2017, 15:55:08 UTC
195f43c Adding reward clipping 04 May 2017, 15:55:08 UTC
bd502fa Initial commit for ReinforcementLearning tutorial 04 May 2017, 15:55:08 UTC
db7ea22 Integrate wilrich/fixReplacePlaceholderTest into master 04 May 2017, 12:02:25 UTC
2f0973b Integrate alrezni/scratch into master 04 May 2017, 09:19:59 UTC
5dbe52d Fix test_replace_placeholders 03 May 2017, 16:00:37 UTC
3255c9c Integrate alrezni/v2_seed_reader into master 03 May 2017, 14:51:04 UTC
74e9f88 Minor usability improvements in device selection/locking 03 May 2017, 14:47:24 UTC
4080919 Add initial seed reader config option 03 May 2017, 12:47:01 UTC
c8cb24f Fixed typos and updated description 03 May 2017, 05:39:26 UTC
back to top